Question

Exhibit L.1 reports the multivariate odds ratios comparing each category to women who never had an induced abortion and had at least one pregnancy. Researchers were able to interview 845 out of 1,011 (83.5 percent) of the eligible cases and 961 out of 1,239 (78 per- cent) of the eligible controls. Of the cases, only 689 (81.5 percent) had complete information on abortion history, compared to 781 (81.3 per- cent) of the eligible controls.

Abortion History	Cases	Controls	Crude OR2	Multivariate OR (95% CI)3
Ever had abortion	210	201		1.5 (1.2–1.9)
1 abortion only	150	142		1.5 (1.1–2.0)
2+ abortions	60	59		1.6 (1.0–2.4)
Age at first abortion
<18	20	15		2.5 (1.1–5.7)
18–19	34	36		1.7 (1.0–3.0)
20–29	115	123		1.3 (1.0–1.7)
30+	41	27		2.1 (1.2–3.5)
Timing of first abortion
Before 1st birth	69	76		1.4 (1.0–2.0)
After 1st birth	74	63		1.5 (1.0–2.2)
Never gave birth	67	62		1.7 (1.2–2.6)
Never had abortion	479	580		––
Never had an abortion and at least one pregnancy1	466	564		––

Note: 1. Estimated from the data. 2. Multivariate OR adjusts for age, family history of breast cancer, religion, age at first pregnancy. 3. Both crude and multivariate odds ratio estimates risk relative to women with a least one pregnancy who never had an induced abortion. Source: Daling et al. (1994).

Calculate the crude odds ratios for each of the abortion history strata in Exhibit L.1. What is the overall increased risk of abortion after adjusting for several covariates?

Answer 1